29 replies to this topic

[rothbauerw]

Someone sent WRD a test table to look at flipper physics (I don't know who and neither does WRD, it was sent anonymously).  I wasn't quite sure what it was supposed to do, but I thought it was a good idea.  So, I modified the table to collect data on various flipper settings to see what impact they have.

 

Setup:

 

For each test, I configure my flipper settings in the editor and script.  I then open the table.  Once running, I can use the arrow keys to select an x-coordinate along the flipper to test.  Pressing the right magna save, a ball is created and rolls down the right flipper.  When the ball reaches the pre-designated x-coordinate, the script automatically triggers the flipper.  When the ball strikes a wall, the y-coordinate of the strike is captured if it hits the vertical wall and the x-coordinate is captured if it hits the horizontal wall.  I also captured the ball velocity (for some data points) and the time it takes the flipper to reach end of stroke.

[bord]

Damn, that is cool. Glad we've got guys like you working on this.

[nFozzy]

Looks cool. Do you have a link? I made the original for converting my elasticity values over to LFHM. It has nothing to do with flippers, so no idea why I decided to name the file FlipperTraj.vpx... Sorry for the confusion.

 

Here's the original if anyone's interested. It drops balls from various heights and compares the bounces between 2 different physics sets defined in the script.

https://www.dropbox....j_HMLF.vpx?dl=0

[rothbauerw]

Here's my first pass results:

 

 

Table updated with link in post below.

Edited by rothbauerw, 03 June 2017 - 05:23 AM.

[rothbauerw]

Observations from the tests:

 

Coil Ramp Up really widens the angles as you near the tip of the flipper, but provides sharper angles near the base of the flipper.  It also creates significant variation on shots triggered at the same point from the middle of the flipper.  Very unpredictable!  After running these tests, I'd be hesitant to use Coil Ramp Up.

 

I was able to obtain very similar results with Low Mass and High Mass flippers in terms of flipper speed and ball velocity, with the exception that the high mass flippers seem to provide sharper angles near the base of the flipper.  See Low Mass, High Friction and High Mass, 4000 Strength tests.

 

Strangely, I didn't not see a huge difference between the flippers with and without the flipper script mod.  From other testing, I have seen very tight backhands perform better.

 

What other tests would folks like to see?

[wrd1972]

Looks cool. Do you have a link? I made the original for converting my elasticity values over to LFHM. It has nothing to do with flippers, so no idea why I decided to name the file FlipperTraj.vpx... Sorry for the confusion.

 

Here's the original if anyone's interested. It drops balls from various heights and compares the bounces between 2 different physics sets defined in the script.

https://www.dropbox....j_HMLF.vpx?dl=0

Ah Ha. Your the "guest".

[rothbauerw]

It would be helpful to have folks document what the expected flipper behavior is. I have my opinion, but I no longer have a real machine to compare to.

I do have some additional results I'll post tonight that show some quirkiness in the flipper physics.

That being said, I'm leaning towards 1.5 mass flippers with strength in the 4200-4500 range. They produce a very similar results to the low mass high friction flippers without the occasional side effects. Higher mass also seem to have better backhands.

Edited by rothbauerw, 31 May 2017 - 05:21 PM.

[fuzzel]

Oh cool please test all kind of settings please. So we can add good flipper physics values to the PhysicsValues.txt file as reference if someone wants to add a modern, mid or old EM style flipper.

[rothbauerw]

I'm going to spend a little time updating the table for others to use as well.  The more people that can test setting, the better.

 

@nFozzy, thanks for the original table.  It sparked some creative thought for me.  I really like the work you're doing with elasticity.  I might ask for some help on flippers too down the line.

[rothbauerw]

I've updated the table to be pretty self explanatory and should allow almost anyone to tune flippers to their liking (or compare results with various settings).  Use the right magna save to test a specific location on the flipper.  Use the left magna save to run through 6 positions across the flipper.  Flipper physics can be set in the editor and most can be adjusted on the fly using various hot keys and the up and down arrows.

 

The table will provide both numbers and a visual representation of the results.

 

I've also added an inlane drop test.  Use the up and down arrows to select the desired height and press "K" to drop a ball.

 

https://www.dropbox....LF 1.1.vpx?dl=0

 

 

Edited by rothbauerw, 03 June 2017 - 03:32 PM.

[nFozzy]

How do you feel about this spread?

 

[rothbauerw]

The spread looks pretty good to me. Two things to consider: 1) velocities are pretty high, the ball is really going to fly around the table; 2) with a high mass ball and lower mass flipper sometimes you see weird behavior when trying to trap the ball or catching a ball racing from the top of the table.

If you like the velocities and you don't see the other two issues (you might not with the strength setting you're using) then these settings might play very nicely.

I've been using a flipper mass of 1.5. To compensate for the higher mass, I've found increasing the return strength to about 0.13 produces a very nice result.

[toxie]

Your last dload link is dead.. Mind to reupload again?

[rothbauerw]

H

Your last dload link is dead.. Mind to reupload again?

You can find the most recent version over at randr's site with instructions. Look under Tutorials.

[nFozzy]

The spread looks pretty good to me. Two things to consider: 1) velocities are pretty high, the ball is really going to fly around the table; 2) with a high mass ball and lower mass flipper sometimes you see weird behavior when trying to trap the ball or catching a ball racing from the top of the table.

If you like the velocities and you don't see the other two issues (you might not with the strength setting you're using) then these settings might play very nicely.

I've been using a flipper mass of 1.5. To compensate for the higher mass, I've found increasing the return strength to about 0.13 produces a very nice result.

Ball and flipper mass are both 1

[rothbauerw]

Then if you're good with the velocities you should be good to go. I was just going by the ball mass shown in the screenshot.

[ClarkKent]

As far as I tested the test table I found one thing not considered: a ball rolling down the inlane. Because a ball rolling down the inlane changes the behavior of the ball, the angles change dramatically.

[wrd1972]

Okay bringing this one back from the dead.

 

I have played a bunch of real PB lately and the VP10 flippers are seriously tweaking me yet once again.

 

In short and IMO, the ball trajectories are still overall a bit to wide. If the entire range of trajectories were shifted more toward the backhand side, it would be more accurate. And when the ball is hit more on the back end side, it definitively lacks velocity.

 

Here is a quick illustration of what I am speaking to. Note the black lines and their position and length (velocity). To me, the overall flipper physics need to shift more this direction:

 

So above said and using Roths very awesome test table, I simply cant get the trajectories to shift toward the backhand side and increase power on the backhand side, in a significant way without causing some major weirdness. There is just some missing link here.  IMO, its a flipper speed which many here have stated many times in the past.

 

Toxi and fuzzel. Seems like I recall a request being made to introduce flipper speed to the flipper physics options. Was this request feasible without blowing things up? Anyone else have any other opinions on this?

[gtxjoe]

Are you testing with your high ball mass flipper tweak script modifications or is this strictly unmodified VPX physics settings?  Probable will play a factor here

[wrd1972]

Yeah I am using 1.7 mass and 50 size. Just for giggles, I will test with normal mass.